Synthetic benchmarks

[ChrisZH]

Do synthetic benchmarks tax the gpu more than actual gameplay? I see temperatures on the benchmark and i am wondering how they would coincide to actual games when comparing temperatures on gpu.

[Guest]

Generally...synthetic benchmarks stress your gpu/cpu more than games do. Synthetic's try to keep your gpu/cpu at 100% the whole time whereas whilst playing games, your cpu and gpu usage will change depending on where you are and what's happening on screen.

Hope that helped!

[Mira Yurizaki]

It depends. Since purely GPU benchmarks don't tax the CPU much, this allows the CPU to focus mostly on getting GPU commands out. Since the GPU is likely going to be at 100%, then it's going to operate at its upper operating temperature. In games, if the CPU is too busy getting GPU commands out to keep the GPU working (i.e., a CPU bottleneck), then the GPU is going to be working less than 100% and this can potentially result in lower temperatures.

[ChrisZH]

6 hours ago, Mira Yurizaki said:

It depends. Since purely GPU benchmarks don't tax the CPU much, this allows the CPU to focus mostly on getting GPU commands out. Since the GPU is likely going to be at 100%, then it's going to operate at its upper operating temperature. In games, if the CPU is too busy getting GPU commands out to keep the GPU working (i.e., a CPU bottleneck), then the GPU is going to be working less than 100% and this can potentially result in lower temperatures.

Hmm alright, thank you. So a gpu stress test should be treated as like the highest the temps should reach?

[Mira Yurizaki]

5 hours ago, ChrisZH said:

Hmm alright, thank you. So a gpu stress test should be treated as like the highest the temps should reach?

More or less. Though it's more like "could" rather than "should"